You should try GPT, I’d be really interested to hear if it works better. (Exclusively using GPT for systems work at $DAYJOB, but compare with opus every couple weeks and GPT consistently gives me better results)
I've been comparing Claude vs Codex using GPT and Claude consistently is better than GPT about reasoning, about writing code, and using the tools as appropriate.
GPT for instance had a lot of issues using git worktrees, and didn't understand how to correctly use it to then merge stuff back into a main branch, vs Claude which seems to do this much more naturally.
GPT also left me with broken tests/code that I had to iterate on manually, Claude is much better about reasoning through code. Primarily Python.
> GPT for instance had a lot of issues using git worktrees, and didn't understand how to correctly use it to then merge stuff back into a main branch, vs Claude which seems to do this much more naturally.
I wonder how much of that is due to the model being somehow better, or the harness having built-in instructions on how to use them.
I've used worktrees with Codex just fine, but I instructed it to use my scripts for setting it up and tearing it down. The scripts also reflinked existing compilation artifacts to speed up compiling and allocated a fresh db instance for it, but then also applied a simple protocol for locking the master repository during merges, so multiple agents wouldn't try to merge at the same time. It has been following those instructions quite well.
The saving grace here is a rewrite of a project with a good test suite is the sweet spot: LLMs are great at translation and do great with verifiable goals.
I agree it’s still mind blowing compared to before times, though.
It costs several times what it would cost a small team of engineers, even assuming you gave the engineers more time to do it. I'm guessing (wildly) this was around 0.5M USD in compute time. You do get the result quicker, though.
> I'm guessing (wildly) this was around 0.5M USD in compute time.
That seems like an especially wild guess. If you take e.g. Opus 4.7 prices, and make the assumption that you are consuming roughly $30 for every million tokens of output (this comes from just summing the $25 per million tokens of output and $5 per million tokens of input and assuming that caching basically makes all that work out), and assume an output rate of 80 tokens per second (which seems like a high estimate based on online searching), it would take you about 2411 days of non-stop Opus 4.7 usage to hit 500k in API spend.
The only way you could possibly run that amount of usage in 6 days is if you were running ~400 instances in parallel. From personal experience, that seems crazy high for this project.
I think you are off by at least an order of magnitude (potentially even 2 depending on how the person is managing agents, but I could see something like dozens of agents 24/7, so I'm way less confident in 2, but I think it's still more likely to be closer to 10-20k in API spend).
But usually companies are much more careful before even spending that half a million. (And most companies don't have that money sitting around.) They would do small PoCs, do comprehensive benchmarks and evaluations of those PoCs, and decide whether to actually go ahead, and, more importantly, stick to it.
Being able to afford half a million doesn't mean you do it on a whim, or just throw all of that away if things don't go well.
But what do I know. I am nothing compared to our AI overlords like Anthropic.
If I was a frontier lab and I solved continual learning, as of today I would absolutely not release it - the society isn't ready for this; society isn't even ready for widespread diffusion of current publicly available frontier models.
If however I was a frontier lab who solved continual learning and my competitor also solved and released it, I would release mine immediately, obviously.
The point is, continual learning might be solved already, we just don't know and those who might know would rather keep their mouths shut. It isn't my base case (financial situation of frontier labs is such that they'd probably release immediately as long as they have inference compute to serve this revolutionary capability), but it isn't impossible.
You're not a frontier lab, the shareholders own those. And if shareholders get a private briefing about an unprecedented breakthrough in continual learning, they would announce it from the rooftops to take credit for the progress ASAP and reap the rewards for their stock value.
The only lab that I can exempt from this is DARPA.
Shareholders are not insiders. Public companies do secret projects all the time of which shareholders know absolutely nothing about and may never learn the details of them if they get cancelled.
No youre missing the point of the poster - disclosures in private markets are different than public, especially in the context of large commitments - the company has no choice.
The implications are very different if everyone owns them even if they aren’t public. They may have no choice whether to share, but the owners which have the privilege to know (not everyone because earlier owners aren’t stupid) don’t act the same, right?
And let’s be honest - rules get bent all the time, especially when valuations are 9 figures. Stakeholders at this point won’t risk killing a golden goose.
What does this have to do with the subject at hand? Is the internet like this now, that in every message board there is just islands of content floating in oceans of snark?
Foreign investment isn't fake growth and money being spent in the country is definitely a good thing. It's how Singapore managed to kickstart its economy in the 1960s. Lee Kuan Yew tried very hard, and succeeded, in getting foreign corporations to set up shop in Singapore. The key is to capture value and move up the chain over time rather than getting stuck as a "cheaper back office".
Yep, and today the situation is completely reversed. Through acquisition and business development Singapore is the country which owns the brands and invests in other countries. Poland just needs to stick to the formula. It's citizens are building global-class professional, managerial, and business development experience. Soon if not already those employees will start itching to build their own businesses. Poland just needs to maintain a competitive environment, and not let international companies suppress local startups by lobbying for anti-competitive laws and policies that favor the big guys, foreign or domestic. If it wants to give local companies a leg up, do it indirectly by investing in education and research.
This is zero sum thinking. The foreign companies benefit and the local Polish people benefit. Wealth is created in the process and everyone benefits. What if those companies never came and never employed Polish people? Would Poland be any better off?
if spotify employs an american and they become more experienced over their tenure were american resources extracted? human capital tends to get better with experience, particularly when dealing with high quality foreign management.
this is Poland for you. everyone complains about everything. perhaps that's the secret to success - there's always something to complain about and one in a hundred (or thousand) people actually does something about it.
> this is Poland for you. everyone complains about everything
That is western Europe for you, not just Poland. Same in the Netherlands, same in Sweden, same in Belgium, same in Denmark, same in Norway, same in France, same in Germany, etcetera. Descartes claimed that he thought, therefore he was. A more realistic and equally erudite quote would be Queror, ergo sum which translates to I complain, therefore I am.
(also, q.e.d. because I'm complaining about people complaining)
The specific anti-EU complaining among the less-educated, older and rural population which benefits a lot from Poland being an EU state is not the same in Netherlands, Sweden, Belgium, Norway.
It is the same in Hungary and Slovakia though, for a reason.
Today's news brings us the following headline:
"Fugitive former Polish Justice Minister Zbigniew Ziobro is now in the United States courtesy of a visa from President Donald Trump after fleeing Hungary, local media report."[1]
In case it is not apparent to you, Ziobro was running on the PiS party ticket — the very same party running on the anti-EU sentiment that the people we're discussing have, in all certainty, voted for.
That's in the wake of another PiS-aligned fugitive judge, Tomasz Szmydt, fleeing investigation for espionage to find refuge in... Belarus[2].
Naturally, he condemned Poland for being too "pro-Western".
Swedish judges don't do that, while hiding in Belarus.
You don't have PiS in Sweden.
And Trump isn't welcoming fugitive Justice Ministers from Sweden into the US because they can't hide in Hungary anymore..
...because the Hungarian MAGA lost the elections there (in spite of JD Vance flying over to campaign for Orban, Orban giving speeches at CPAC, and Trump explicitly praising him).
Do you realize how off mark your objection to applying "the MAGA moniker" to PiS supporters is?
It's not even metaphorical; they're directly collaborating on a political level.
Are they really unaware or do they actively deny the cultural connections? Prussia was a thing not that long ago - it is still used as a slang term for Germans in parts of the Netherlands ("de Pruisen" of "die Preußen"). Anyone who had a bit of history or who has looked at an older map sees that Prussia was divided between what is now Germany and Poland. Of course both countries went through a lot of upheaval between then and now but there's still plenty of people alive who will remember living in Prussia.
the kubernetes reveal had me literally in tears
reply